New solution for fully automated analysis and reporting of routine musculoskeletal X-rays

Project Information

AutoRay

Grant agreement ID: 855101

DOI

10.3030/855101

Project closed

EC signature date 18 February 2019

Start date 1 March 2019

End date 31 May 2019

Funded under

• INDUSTRIAL LEADERSHIP - Innovation In SMEs

Total cost

€ 71 429,00

EU contribution

€ 50 000,00

Coordinated by

RADIOBOTICS APS
Denmark

Periodic Reporting for period 1 - AutoRay (New solution for fully automated analysis and reporting of routine musculoskeletal X-rays)

Reporting period: 2019-03-01 to 2019-05-31

Summary of the context and overall objectives of the project

Clinical imaging is essential to support diagnosis and therapeutic decision-making. Plain x-rays remain the main imaging technique for routine and emergency examinations. The length of time to analyze and report, and the potential risk to patients on receiving delayed treatments is a matter of concern in multiple countries. The flux of x-rays collected is superior to the capacity of the radiologists to analyze and prepare the medical reports in a timely manner, creating large backlogs.
Currently, the number of radiologists is insufficient to fill the demand. According to a UK workforce census 2017 report from the Royal College of Radiology, the diagnostic reporting workload has increased 30% in the last 5 years and that there is a shortfall of more than 1000 radiology consultants (open positions)
The objectives of the project are to develop a novel solution to automatically analyze and generate reports describing medical images. The software uses machine learning, combined with advanced image processing (computer vision) to accurately analyze x-ray images of the msk system in a standardized manner.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

We have interviewed 12 radiologists and medical doctors in 4 countries to document the requirements and technical needs to accept a solution to generate reports describing medical images of msk x-ray. We have quantified the addressable market and developed a business model. On the regulatory side, we have investigated certification requirements.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

We have developed a prototype of automatic report where clinical findings are used to generate automatic report describing clinical findings and impressions. Uses in two hospitals have tested it and gave feedback on its performance in term of feature detection and reporting accuracy. The software has been integrated in in open source APCS server, which has enabled us to perform integration tests and usability in-house.
An automatic report will enable patients to receive diagnosis, and thus potentially treatment, from their treating physician much faster.

image-radiobotics.png